Case Study: Leveraging Generative AI for Community Health Insights in Canada

Introduction

In Canada, community health programs often rely on field workers who interact directly with populations, collecting invaluable observations about community concerns, trends, and needs. These interactions produce a wealth of unstructured data, yet analyzing it effectively for actionable insights has historically posed challenges. Auxiliobits stepped in to streamline and elevate this data analysis process, enabling the organization to make data-driven decisions more effectively and precisely.

Challenge

Field workers collected observations in unstructured formats, logged within SharePoint lists and other structured databases. However, with thousands of entries, recognizing trends and patterns—especially those nuanced by socio-demographic details—was arduous and time-consuming when handled manually.

These obstacles prevented healthcare providers from anticipating community needs in real-time, potentially delaying response measures for emerging health concerns. The organization needed an AI-driven solution that could process vast, unstructured data volumes, identify hidden patterns, and yield insightful visualizations for decision-makers.

Solution

Auxiliobits deployed a generative AI and data analytics solution tailored to handle both structured and unstructured data. This solution transformed the organization’s approach to data, facilitating the discovery of actionable insights and enabling deeper understanding of population health trends.

The solution’s functionality and technical details were as follows:

Functional Aspect

Auxiliobits’ solution extracted trends from field observations in SharePoint and other large datasets, capturing insights critical to the community health program. By employing generative AI, it became possible to analyze unstructured data entries to reveal hidden patterns across demographics, assisting in more targeted health interventions.

Dynamic Data Analysis: Field data, including unstructured observations, was analyzed for socio-demographic trends, allowing insights into age, location, health behaviors, and recurring health needs.
Real-Time Insights: The organization gained timely insights that could drive strategic planning and help adjust health interventions according to emerging community needs.
Scalable to Multiple Datasets: Beyond SharePoint, the solution’s versatility allowed it to ingest and analyze large data sets from various sources, enabling a more comprehensive view.

Technical Architecture

Data Extraction from SharePoint: Unstructured data from SharePoint lists—encompassing field notes, observations, and assessments—was extracted using APIs to retain all metadata and relationships. These aspects were crucial for contextual and semantic analysis.
Ingestion and Vectorization: Data was ingested into a vector database, such as Azure Cosmos DB, where it was processed and embedded into vector representations, enabling efficient and semantic-based search.
Semantic Retrieval via Azure AI Search: Azure AI Search used vector embeddings for advanced, context-sensitive data retrieval, ensuring that retrieved data was relevant to healthcare providers’ queries.
LLM for Data Analysis: A large language model (LLM) analyzed retrieved data, identifying key IDs, extracting relevant insights, and recognizing trends. For example, recurring themes in field observations were automatically categorized, supporting the program in timely identification of new health concerns.
Power BI Visualization: Refined data was visualized through Power BI dashboards, offering leadership insights into demographic patterns and actionable health trends. This enabled the community health program to proactively respond to needs across different regions.

Results and Impact

Improved Decision-Making: The program gained real-time insights into community health trends across demographics, enabling faster and more strategic response to emerging needs.
Significant Time Savings: By automating data analysis, the program reduced hours previously spent on manual data processing, allowing team members to focus on high-priority initiatives.
Enhanced Health Outcomes: With access to timely, actionable insights, the organization was better positioned to address the needs of the community in a proactive manner.

Conclusion

Auxiliobits’ generative AI solution demonstrated the potential of advanced AI-driven analytics in transforming unstructured data into valuable insights. This solution empowered the Canadian community health program to leverage data in ways previously unattainable, ultimately improving service delivery and response times. Through its expertise in Agentic Process Automation, Auxiliobits remains committed to helping organizations extract actionable insights from complex data sets, ensuring smarter, data-driven decision-making across sectors.